Six months later, we revisited Realtek's drivers and found that little had changed. With the then-current 1.66 driver release, EAX occlusions and obstructions still weren't being implemented properly, if at all. Again, we contacted Realtek, but the crab remained silent.
I didn't expect to cover the issue again, particularly because Windows Vista has dropped support for hardware-accelerated DirectSound 3D, requiring special software that thus far only Creative provides for its own audio products. However, last week a reader emailed me suggesting that Realtek had fixed occlusion and obstruction support in its latest 1.71 drivers, so I decided to take them for a spin.
With its 1.71 drivers, Realtek appears to have correctly implemented EAX occlusions and obstructions, at least in RightMark 3D Sound's positioning accuracy test. There, sounds are properly muffled behind obstruction and occlusions planes that had no effect with Realtek's previous drivers. And there was much rejoicing—until I fired up Battlefield 2. This game had been unplayable with older Realtek drivers and EAX effects enabled, where the lack of proper occlusion and obstruction support resulted in the player being bombarded by sounds from all over the battlefield. Unfortunately, little has changed with the 1.71 drivers; you're still assaulted by sounds that should be muffled by buildings, walls, and other in-game elements.
So while Realtek's latest drivers may correctly implement EAX occlusion and obstruction effects in RightMark 3D Sound's positioning accuracy test, that implementation doesn't appear to translate to games. Interestingly, though, the new drivers do consume more CPU cycles in RightMark 3D Sound's EAX benchmark.
The difference in CPU utilization between the older 1.66 drivers and the 1.71s is only a couple of percentage points at 16 buffers, but it grows as the number of buffers increases. Realtek's latest 1.72 drivers also exhibit this higher CPU utilization. Although they correctly implement occlusion and obstruction effects in RightMark's positional accuracy test, Battlefield 2 remains a mess.
The world of integrated motherboard audio is entirely too mundane for conspiracy theories, but it's odd that Realtek's latest drivers support EAX occlusion and obstruction in a benchmark—and with a hit to CPU utilization—but not in a game. Competing solutions from Analog Devices have no problems with either.
|1. Hdfisise - $600||2. Ryszard - $503||3. Andrew Lauritzen - $502|
|4. the - $306||5. SomeOtherGeek - $300||6. Ryu Connor - $250|
|7. doubtful500 - $200||8. Anonymous Gerbil - $150||9. webkido13 - $135|
|10. cygnus1 - $126|
|ASRock serves up a new pair of Kabini embedded mobos||4|
|Samsung Electronics' Q2 guidance shows falling revenues, profits||9|
|AMD issues updated statement on Fury X noise problems||51|
|AMD revises Q2 guidance; gross margin, revenue fall||55|
|TSMC's 10-nm FinFET process toddles towards validation||21|
|Samsung cranks SSDs to 2TB with the 850 Pro and EVO||69|
|TR's July 2015 mobile staff picks||39|
|Minecraft: Story Mode gets a trailer, more details at Minecon||11|